Can morphology reliably distinguish between the copepods Calanus finmarchicus and C. glacialis, or is DNA the only way?

(1)

Can morphology reliably distinguish between the copepods Calanus finmarchicus and C. glacialis, or is DNA the only way?

Marvin Choquet ,

¹

* Ksenia Kosobokova,

²

Sławomir Kwa sniewski,

³

Maja Hatlebakk,

^1,4

Anusha K. S. Dhanasiri,

¹

Webjørn Melle,

⁵

Malin Daase,

⁶

Camilla Svensen,

⁶

Janne E. Søreide,

⁴

Galice Hoarau

¹

1Faculty of Biosciences and Aquaculture, Nord University, Bodø, Norway

2P.P. Shirshov Institute of Oceanology, Russian Academy of Sciences, Moscow, Russia

3Institute of Oceanology, Polish Academy of Sciences, Sopot, Poland

4Department of Arctic Biology, The University Centre in Svalbard, Longyearbyen, Norway

5Institute of Marine Research, Bergen, Norway

6Faculty of Biosciences, Fisheries and Economics, Department of Arctic and Marine Biology, UiT The Arctic University of Norway, Tromsø, Norway

Abstract

Copepods of the genusCalanusplay a key role in marine food webs as consumers of primary producers and as prey for many commercially important marine species. Within the genus,Calanus glacialisandCalanus finmarchi- cusare considered indicator species for Arctic and Atlantic waters, respectively, and changes in their distributions are frequently used as a tool to track climate change effects in the marine ecosystems of the northern hemisphere.

Despite the extensive literature available, discrimination between these two species remains challenging. Using genetically identified individuals, we simultaneously checked the morphological characters in use forC. glacialis andC. finmarchicusidentification to compare the results of molecular and morphological identification. We studied the prosome length (1); the antennules and the genital somite pigmentation (2); the morphology of the fifth pair of swimming legs and of the mandible (3). Our results show that none of these morphological criteria can reliably distinguish betweenC. glacialisandC. finmarchicus. This has severe implications for our current understanding of plankton ecology as a large part of our knowledge ofCalanusmay be biased due to species misidentification and may subsequently require reinvestigation with the systematic use of molecular tools.

Copepods of the genusCalanusare the dominant compo- nent of the zooplankton in the North Atlantic and the Arctic (Jaschnov 1972; Fleminger and Hulsemann 1977; Conover 1988; Kosobokova et al. 2011; Kosobokova 2012) and are by far the most studied zooplankton species, with ca. 100 scientific publications per year for the last 30 years (Web of Sci- ence). They play a key role in marine food webs as consumers of primary producers and microzooplankton and as prey for many commercially and non-commercially important species (Gislason and Astthorsson 2002; Beau- grand et al. 2003; Skjoldal 2004; Varpe et al. 2005; Michaud and Taggart 2007; Steen et al. 2007; Falk-Petersen et al.

2009). Furthermore, they are key drivers of the vertical export of material from the upper part of the water column due to the ability of packing organic material into large fast- sinking fecal pellets (Wilson et al. 2008). In marine food webs,Calanusspp. are essential agents of matter and energy transfer between phyto- and microzooplankton and higher trophic levels.

In the North Atlantic and Arctic regions, the Arctic species Calanus glacialisand the smaller north Atlantic Calanus finmarchicus account for most of the zooplankton biomass (Fleminger and Hulsemann 1977; Hassel 1986; Blachowiak- Samolyk 2008; Søreide et al. 2008; Kosobokova and Hirche 2009; Kosobokova 2012). The spatial distribution of these two copepods is linked to the distribution of Arctic and Atlantic waters, respectively, and they are thus considered indicator species for these water masses (Jaschnov 1966;

Jaschnov 1970; Unstad and Tande 1991; Bonnet and Frid 2004; Daase and Eiane 2007; Helaou€et and Beaugrand 2007;

Blachowiak-Samolyk 2008; Broms et al. 2009). Recently, C.

glacialis and C. finmarchicus have been regarded as beacons

*Correspondence: [email protected]

Additional Supporting Information may be found in the online version of this article.

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

LIMNOLOGY

and

OCEANOGRAPHY: METHODS

Limnol. Oceanogr.: Methods16, 2018, 237–252 VC2018 The Authors Limnology and Oceanography: Methods published by Wiley Periodicals, Inc. on behalf of Association for the Sciences of Limnology and Oceanography doi: 10.1002/lom3.10240

(2)

of climate change (Hays et al. 2005; Wassmann et al. 2015), as changes in their distribution are interpreted as changes in Atlantic water circulation and potential “Atlantification” of the Arctic (Wassmann et al. 2006; Falk-Petersen et al. 2007).

The ecological importance ofC. finmarchicusandC. glacialis is unquestionable, but distinguishing between them in regions of co-occurrence has always been challenging (Unstad and Tande 1991; Hirche et al. 1994). Three main morphological characters have been used, (1) prosome length; (2) redness of antennules and genital somite (the two spermathecae); (3) structure of the fifth pair of swimming legs and the coxal endid of the mandible (in adults).

Because of convenience, the prosome length measurements (1) have been and remain the most commonly used method to separate the two species (see, for example: Unstad and Tande 1991; Kwasniewski et al. 2003; Arnkværn et al.

2005; Forest et al. 2011; Hirche and Kosobokova 2011; Koso- bokova 2012) although several recent studies have demon- strated a size-overlap in specific regions (Lindeque et al.

2006; Parent et al. 2011; Gabrielsen et al. 2012).

Another trait that has been recently suggested to distinguish betweenC. finmarchicusandC. glacialisis the presence or absence of red pigmentation on their antennules and, in the case of adult females, on their genital somite (originally genital field) (2) (Nielsen et al. 2014). Examination of this character requires that individuals are alive, so the samples have to be sorted directly after collection, which is also a challenge.

The classical, but most complex and time-consuming approach to identify C. finmarchicus and C. glacialis is to examine their morphological characters (3) that have been suggested as diagnostic of the two species. Most common is the examination of the structure of the fifth pair of swimming legs in adult females and males (Jaschnov 1955), and the morphology of the coxal endid of the mandible (gnathobase) (Beklemishev 1959). Examination of both characters requires performing a fastidious and specific preparation on each specimen, and is therefore seldom applied during routine zooplankton samples analyses.

Although several diagnostic molecular markers have been developed for Calanus, from mtDNA RFLP (Lindeque et al.

1999) to nuclear InDels (Smolina et al. 2014), their use in the zooplankton research community has so far remained limited.

A recent reappraisal ofCalanusspp. distribution in the North Atlantic/Arctic Oceans relying on large scale sampling and molecular identification has suggested that misidentification is widespread and has led to erroneous conclusions regarding Calanusbiogeography (Choquet et al. 2017).

Species misidentification may be less problematic in studies focusing on describing zooplankton assemblages based on higher taxonomic categories (e.g., Aßmus et al. 2009) or in trait-based studies, which aim at investigating ecological func- tions of assemblages (e.g., Brun et al. 2016). A correct species identification is however crucial for understanding species- specific life history strategies, species-specific productivity

estimates and for studying distribution patterns, particularly if species are considered indicative for specific water masses and if changes in their distribution are assumed to have far reach- ing ecosystem impacts.

Both species differ in life strategies such as energy require- ments for reproduction and growth, timing of reproduction, composition of overwintering populations, and seasonal vertical migration patterns. These differences reflect adaptations to the environmental conditions in their main areas of distribution (Falk-Petersen et al. 2009), with C. glacialishaving adapted more flexible life history strategy to deal with the constrains of seasonally ice-covered seas (Daase et al. 2013) and low temperature leading to a larger body size and longer life span compared to C. finmarchicus. It is crucial to cor- rectly identify them to understand their life history adaptations fully, how they have evolved differently in each species and how climate change will be affecting each species’ productivity, population success, distribution and role in the food web. Using prosome length to discriminate between species has shown to underestimate smaller sizedC.

glacialis (Gabrielsen et al. 2012), which may bias species- specific biomass estimates and our understanding of energy allocations in that species.

In the present study, we use molecular tools to assess the reliability of the morphological characters used to discriminate between C. finmarchicus and C. glacialis across a large part of their distributional range.

Material and procedures

Samples collection and pre-sorting

Zooplankton were sampled in fjords along the Norwegian coast, in the White Sea, in Svalbard waters and in the Nansen Basin (Table 1) by vertically towed plankton nets (WP-2/Juday types) with mesh sizes between 150 lm and 200 lm. The whole water column was sampled for most of the locations, except for the White Sea (100–0 m) and the Svalbard fjords (20–0 m). The sampling locations were selected to represent a latitudinal gradient from the southernmost (Lurefjord) to the northernmost (Nansen Basin) areas of co-occurring ofC. fin- marchicusandC. glacialis. The White Sea, where onlyC. glacialis occurrence was reported historically (Jaschnov 1955;

Jaschnov 1966) and recently confirmed genetically (Choquet et al. 2017), and the region of Raunefjord/Korsfjord where onlyC. finmarchicusoccurrence was reported, were also sampled in order to have more elements of comparison. Directly after sampling, a Folsom plankton splitter was used to randomly subsample100–200 live individuals of the older (CIV) copepodite stages. Prosome length measurements and examination of the redness of antennules and genital somite (for details, see below) were carried out right after sampling, on the subsampled individuals kept alive in seawater. These live individuals were subsequently preserved individually in

(3)

70–80% undenatured ethanol for later molecular-based species identification and morphological examinations.

Prosome length measurements

We subsampled up to 200 live individuals of late copepodite stages IV, V, and CVI female (CIV, CV, and CVI F) ofCalanus per sample from each of the nine locations (pooled into six geographically distant regions—895 specimens in total) (Table 1). For the sampling locations where it was possible, photographs of individuals were taken with a camera attached to a stereomicroscope. The prosome length of each specimen was measured from the tip of the cephalosome to the distal lateral end of the last thoracic somite (Fig. 1) either using the ruler in the eye-piece of a stereomicroscope to measure directly (resolution 1 lm), or by using cellSens Standard software (version 1.8.1—Olympus corporation^V^C 2009–2013) to analyze the

photograph taken (resolution 0.01lm). All the 895 individuals were identified with molecular markers (see “Molecular species identification” section below). Correlation between latitude and body size (prosome length) was tested independently for C. finmarchicusandC. glacialis, and separately for each developmental stage (CIV, CV, and CVI female) with use of Pear- son’s correlation (in Microsoft^V^RExcel^V^Rversion 14.7.3).

Redness assessment

We evaluated the potential of red pigmentation (“redness”) on antennules and genital somite to separate liveC. finmarchi- cusandC. glacialis, as suggested in Nielsen et al. (2014). A total of 903Calanusindividuals of developmental stage CIV to CVI (adult female and male) from six distant populations in the North Atlantic and Arctic Oceans were investigated in regard to their antennule redness (Table 1). Additionally, pigmentation of the two spermathecae on the ventral surface of the genital somite (the first urosome somite) in adult females was examined for 168 individuals from the same populations. All the individuals examined for their redness were subsequently identified with molecular markers (see “Molecular species identification” section below).

The degree of antennule red pigmentation (“redness”) was very heterogeneous among the studied individuals. We distin- guished four different categories of individuals: antennules with more than 90% of redness; from 50% to 90% of redness;

from 10% to 50% of redness; and less than 10% of redness. The percentage of redness used to distinguish different categories is based on the subjective evaluation of how much of the surface of antennules is red, and how dense this pigmentation is (see Fig. 2 for examples of each category). This choice is justified by our search for a parameter that could be easily and quickly used for routine species identification especially in the field. Objec- tive quantification of redness using an image analysis software Table 1. Sampling locations with positions, sampling dates, and number of individuals used for each analysis.

N ind. analyzed

Location GPS Date Redness

PL Ant Gen Legs/Gnath

Nansen Basin 87800⁰N 55847⁰E 10/4/16 96 94 0 0

Svalbard Isfjord 78819⁰N 15809⁰E 6/5/16 136 227 60 0

Van Mijenfj. (VM) 77846⁰N 15802⁰E 6/3/16 90 0 0 16

White Sea 66833⁰N 33843⁰E 8/22/16 116 115 1 0

Sørfolda (Sorf) 67835⁰N 14850⁰E 4/20/16 0 0 0 7

Salten/Skjerstadfj. Saltenfjord (SALT) 67816⁰N 14838⁰E 2/15/16 72 190 102 24

Skjerstadfj. (SKJ) 67815⁰N 14850⁰E 7/12/16 109 0 0 2

Lurefjord (Lure) 60841⁰N 05809⁰E 6/22/16 188 189 5 22

Raune/Korsfj. Raunefjord 60817⁰N 05808⁰E 6/4/16 43 88 0 0

Korsfjord 60811⁰N 05812⁰E 6/6/16 45 0 0 0

Arctic locations are presented first, starting with the northernmost; the Atlantic locations are listed from North to South. Number of individuals analyzed is given (“N ind. analyzed”), with the precision for the three different analyses: PL, prosome length measurements; Redness—Ant/Gen, examination of redness of antennules/genital somite; Legs/Gnath, examination of morphology of the 5^thpair of legs and mandibular gnathobase.

Fig. 1.Calanussp. adult female, structure of the body.

Choquet et al. Morphological misidentification in Calanus

(4)

(as in Nielsen et al. 2014) was not performed, as the sorting of live individuals was mostly done in conditions precluding taking high-quality photographs, required for such analysis.

Statistical differences in antennule redness between the two speciesC. finmarchicusandC. glacialis, among the different developmental stages, and among the locations sampled were tested using the Kruskal-WallisHtest.

To evaluate the pigmentation of the genital somite, we considered only two categories: red or pale (Fig. 3). Individuals with any redness on one or two of the spermathecae were assigned to “red”; the individuals for which no redness at all on the genital somite was noticeable were reported as “pale.”

Molecular species identification

Each Calanus individual used for this study was genetically identified (913 individuals in total—Table 1). Molecular species identification followed the procedure described in Choquet et al. (2017). In brief, DNA was extracted from ani- mals’ antennules using the HotSHOT DNA extraction method (Montero-Pau et al. 2008) and six nuclear molecular markers of the type InDels (Insertion or Deletion motifs—

Smolina et al. 2014) were amplified by Polymerase Chain Reaction (PCR). PCR amplicons were sized using a 3500xL Genetic Analyzer (Applied Biosystems, U.S.A.), generating a species specific profile (Smolina et al. 2014). Together, these six markers allow the reliable identification of Calanus species in the North Atlantic and Arctic Oceans (Nielsen et al.

2014; Smolina et al. 2014; Choquet et al. 2017). This method

allows the genotyping of each individual for species identification without using or destroying the animal’s body. Once the antennules are removed, the rest of the body is intact and can still be examined for morphology.

Fifth pair of legs and gnathobase morphology examination

Seventy-one individuals from five different locations (Table 1) were examined (49 individuals of developmental stage CV and 22 individuals of CVI adult females), by following a specific procedure. M. Choquet selected the individuals among the genetically identified specimens preserved in ethanol, in order to have both species represented. The 71 selected ones were sent to S. Kwasniewski for dissection (see procedure below), without giving any information about the molecular results of species ID for these particular individuals. After dissection, photographs of the dissected body parts were taken for each individual by S. Kwasniewski, and shared with K. Kosobokova. Examination of the fifth thoracic leg (swimming leg P5—Figs. 4a,b, 5a,b) and the coxal endid of the mandible (gnathobase—Figs. 4c,d, 5c,d) were carried out by both S. Kwasniewski and K. Kosobokova independently, based on the photographs only. Their species identification decisions, based on the pictures analysis, were then sent back to M. Choquet to compare with molecular results. We decided to follow this approach in order to avoid any bias in the expert interpretation of the pictures due to the prior knowledge of molecular ID.

Fig. 2.Categories of red pigmentation of antennules inCalanus. Four photos are shown as examples of the four categories defined as follows: less than 10% of red pigmentation; between 10% and 50% pigmentation; between 50% and 90% of pigmentation; and more than 90% pigmentation.

Fig. 3.Pigmentation of genital somite inCalanus. Pigmentation is defined as red (left photo) or pale (right photo).

(5)

For the examination of the P5 morphology, descriptions of the leg structure provided in Jaschnov (1955); Frost (1974);

and Brodskii et al. (1983) were used. The P5 inCalanuscon- sists of a remnant of precoxa, well developed coxa (basipod 1) and basis (basipod 2), from which two 3-segmented rami (exo- pod and endopod) grow out (Huys and Boxshall 1991). Exami- nation focused on the lamellar structure with denticulated edge, termed also the denticulated lamella, which is present on each of the two coxa (basipod 1) of P5. The denticulated lamella extends along the medial margin of the coxa from the intercoxal plate to nearly the distal medial corner of the segment. According to the species morphological descriptions (Jaschnov 1955; Frost 1974), inC. finmarchicus, this lamella is straight or almost straight, missing clearly expressed incurva- tion characteristic forC. glacialis(seeFig. 4a,b). InC. glacialis, the denticulated lamella is clearly concaved, with well- expressed curvature (deflection) slightly shifted to the posterior surface of the segment in its middle part (seeFig. 5a,b).

For examination of the gnathobase, descriptions provided in Beklemishev (1959), Vyshkvartzeva (1972), and Vyshkvart- zeva (1976) were used. The gnathobase is the coxal endid (a medially directed process on the protopodal segment of the appendage), bearing the toothed cutting edge distally (Huys and Boxshall 1991). The cutting (masticatory) edge of the gnathobase bears several groups of teeth varying in form and structure. Some of these teeth (at least in sexually developed stages) are covered with silicate crowns. In adult females ofC.

glacialis and C. finmarchicus, the complete arrangement of

gnathobase cutting edge includes ventral (V1–V2), central (C1– C₄), and distal (D₁–D₃) teeth plus flexible setae with one or two rows of spines. Between groups of V and C teeth, there is a diastema (a gap between the teeth). Tooth V₂does not have a crown and teeth of group D are often equipped on their lateral surfaces with small denticles. According to Beklemishev (1959) and Vyshkvartzeva (1972, 1976), species-specific differences in the form and arrangement of the teeth concern teeth V₁and V₂. InC. glacialisadult females, the crown of the tooth V1is not very high, compressed in the anterior-posterior direc- tion, and has 2–3 peaks. The tooth V2, which does not have a crown, is well developed and placed on wide cuticular plat- form. Its size is close to the size of V₁and it approximately equals the diameter of its base (Fig. 5c,d). In C. finmarchicus adult females, the tooth V₂is smaller than V₁and its height is larger than the diameter of its base, but its form and size varies (Fig. 4c,d). In comparison withC. glacialis, the tooth V2inC.

finmarchicuspresents as not completely formed.

The examination of the two structures was done after dissection and slide preparation. Each individual from the study collection was first immersed for 10 min in a drop of glycerol : ethanol 1 : 1 mixture placed on a microscope slide with cavity. In 10 min, each individual was photographed using Olympus SC50 CMOS Color Camera, mounted with a photo adapter U-TV0.5xc-3 on Olympus SZX12 Research Ste- reomicroscope, equipped with AXH1x and DFPL2x-3 objectives. The acquisition of the digital pictures was made with Olympus cellSense Imaging Software v.1.12. The pictures of Fig. 4.Morphology of the fifth thoracic pair of legs and the gnathobase of an adult femaleC. finmarchicus(genetically confirmed) exhibiting the species- specific traits as described in the literature. Specimen collected from Van Mijenfjord (ID: VM41). (a,b) Anterior view of the fifth thoracic pair of legs (P5) with denticulated lamellae on the medial margin of the coxa, showing typical “straight form.” Abbreviations: b, basis; c, coxa; en1–en3, endopods 1 to 3;

ex1–ex3, exopods 1 to 3. (c,d) Anterior view of mandible gnathobases, with a typical small second ventral tooth on the cutting edge. Abbreviations: gnth le, left mandible gnathobase; gnth ri, right mandible gnathobase; V1, first ventral tooth; V2, second ventral tooth.

(6)

the body habitus of each individual were made at 10X total magnification, one picture with the use of an AXH1x objective and one with the use of a DFPL2x-3 objective. Then the two structures under consideration were dissected from the body. The P5 was cut off the thoracic somite and placed in a drop of the same glycerol : ethanol 1 : 1 mixture, on a regu- lar microscope slide, anterior side upward.

The mandibles were also dissected one by one from the cephalosome. After removal of the mandible, the gnathobase was dissected from the appendage, and mounted in another drop of glycerol : ethanol 1 : 1 mixture, anterior side upward.

The same procedure was repeated for the second mandible, and finally the pair of gnathobases belonging to one individual was covered with a glass coverslip. The dissection of the appendages and preparation of the microscope slides was done with use of Olympus SZX12 stereomicroscope, at magni- fications ranging from 7X to 90X. In the following step, the investigated structures were photographed using Olympus SC30 CMOS Color Camera, mounted with a photo adapter U- TV1x-2 on Olympus BX51 system microscope, equipped with PlanN 4X and UPlanFLN 10X objectives. The acquisition of the digital pictures was made with use of Olympus cellB Imag- ing Software v.3.3.

Assessment

Prosome length measurements

Based on prosome length measurements of 895 genetically identified individuals from six regions, we confirm that

this character shows a global overlap of size between C. finmarchicus and C. glacialis regardless of developmental stage (Fig. 6). Size frequency distributions, however, differed among different regions. In the Norwegian fjords (Saltenf- jord/Skjerstadfjord; Lurefjord),C. glacialisshowed a complete size overlap with C. finmarchicus, but these C. glacialis were significantly smaller (t-test,p<0.01) than theC. glacialiscap- tured in the White Sea and the high Arctic. Noteworthy, our data showed positive correlations between latitude and body size for bothC. glacialisandC. finmarchicus(Table 2).

Thus, the prosome length cannot reliably discriminate betweenC. finmarchicusandC. glacialisin any of the investigated regions, and even less in the Norwegian fjords. However, in the Nansen Basin and Svalbard waters, the majority of the length values for C. finmarchicus and C. glacialis follow a dichotomy. Prosome length could therefore be used in those particular areas to approximate the overallC. glacialisandC.

finmarchicuscomposition. It has to be kept in mind, however, the inaccuracy of the method leading to underestimation ofC.

glacialis (especially small-sized individuals), and over- estimation ofC. finmarchicusnumbers (Gabrielsen et al. 2012).

Redness assessment

We tested if redness can be used to reliably separate between live C. finmarchicus and C. glacialis, at different developmental stages and across different regions of co- occurrence. According to Nielsen et al. (2014), the genital somite (originally genital field) and the antennules of C.

Fig. 5.Morphology of the fifth thoracic pair of legs and the gnathobase of an adult femaleC. glacialis(genetically confirmed) exhibiting the species- specific traits as described in literature. Specimen collected from Skjerstadjord (ID: SKJ25). (a,b) Anterior view of the fifth thoracic pair of legs (P5) showing denticulated lamellae on the medial margin of the coxa in a typical concave form, with well-expressed curvature. (c,d) Anterior view of mandible gnathobases with the cutting edge with a typical large second ventral tooth on a wide basis (seeFig. 4 legend for abbreviations meaning).

(7)

glacialis adult females had red pigmentation, while the pigmentation of female C. finmarchicuswere mostly pale. How- ever, the study focused only on adult females from a limited geographic location (Greenland) (Nielsen et al. 2014).

In our study, red pigmentation of antennules was variable (Figs. 7, 8, Supporting Information Fig. 2a,b), with significant

differences in redness between the two species, among developmental stages, and locations sampled (Supporting Information Table 1). Antennule redness was assessed for 903 individuals, from copepodite stage CIV to adult females and males, at six different locations (Table 3a–f; Fig. 7). Molecular identification of these 903 individuals was performed consecutively.

Fig. 6.Stage-specific length frequency distributions of prosome length (mm) for copepodites CIV, CV, and adult females ofC. glacialisandC. fin- marchicusin different regions. In total, 895 individuals were measured, from nine locations, pooled into six distant regions, in the North Atlantic and Arctic Oceans. OnlyC. glacialis(blue) occurred in the White Sea, and onlyC. finmarchicus(red) occurred in Raunefjord/Korsfjord area.

(8)

At the northernmost location, the Nansen Basin (Table 3a), all genetically identified C. glacialis had>10% redness on their antennules, and they were all adult females. Stages CV and adult females of genetically identifiedC. finmarchicus individuals collected in the same place were mainly pale except for three females with a slight redness (10–50%).

In Svalbard, the majority of C. glacialis identified genetically, including stages CIV, CV, and adult females, had also>10% redness (except one CV and one adult female with<10%—Table 3b).C. finmarchicusindividuals, including stages CIV, CV, and CVI, tended to be paler compared to C.

glacialis in Svalbard, but some C. finmarchicus, especially females, had>10% redness. One male was detected there, identified asC. finmarchicuswith pale antennules.

Only C. glacialis was detected in the White Sea sample (stages CIV, CV, and adult females—Table 3c). Individuals from stage CIV exhibited almost none, or very little (less than 50%) redness, but a stronger red pigmentation was observed for the older stages CV and adult females.

In the boreal fjords Saltenfjord and Skjerstadfjord (Table 3d), C. glacialis individuals (stages CIV–CVI) most often (88%) had red pigments. All males, 4% of the females and 20% of the CVC. glacialis were pale. The majority of C. finmarchicusindividuals were pale in these two fjords, independently of the developmental stage, however, with three exceptions (one CIV and two adult females). Interestingly, males of both species were totally pale.

In Lurefjord (southern Norway—Table 3e), the majority (73%) of C. glacialis had red pigmentation, with 25% pale CV and 100% pale females. In comparison, the majority (84%) ofC. finmarchicus(CVs) were pale there.

In the open southern fjords Raunefjord and Korsfjord (Table 3f), we only identified C. finmarchicus among the older stages (CIV and CV) in our samples and 56% of these were pale and another 16% had 10–50% redness.

Despite identifying significant differences (Kruskal-Wallis Htest) in the redness of antennules between species, among stages for each species, and among locations for each species for every set of variables compared (Supporting Information Table 1), the general trend was that the majority of individuals of C. finmarchicus tends to have pale antennas whereas the majority ofC. glacialistends to have red ones (Figs. 7, 8).

In both species, there were exceptions, especially for C.

glacialisin the White Sea andC. finmarchicus in Raunefjord/

Korsfjord (Supporting Information Fig. 2a). The tendencies in pigmentation were similar for the different developmental stages (Supporting Information Fig. 2b) except that males of both species were pale without exception (albeit only a few males were investigated) andC. glacialisCIV in general being less pigmented than C. glacialis CV and adult females.

Antennule redness thus appears not to be a reliable diagnostic feature and is clearly not a species-specific trait. It was never 100% diagnostic for any of the six regions investigated. Assessment of pigmentation might be useful to get an overall impression of the species composition in the Arctic Ocean and in isolated fjords, taking into account the error threshold (region dependent), and the fact that investigations have to be done on live organisms.

Regarding the redness of the genital somite of Calanus females, all the C. finmarchicus examined had pale spermathecae, although we only found females of this species in Svalbard and Saltenfjord/Skjerstadfjord (Table 4). Most of the C. glacialis examined (from four regions) had red genital somite, but also a few individuals had pale genital somite in each region. Importantly, we noticed that all the individuals with red genital somite wereC. glacialis. Furthermore, all the individuals displaying both red antennules and red genital somite were C. glacialis(Table 5). Although our results indi- cate that redness of the genital somite is also not 100% diagnostic for species identification, this character seems to be useful to get a global idea of species composition of a zooplankton sample, but using it may result in an underestimation ofC. glacialisnumber of individuals.

Fifth pair of legs and gnathobase morphology examination

The curvature of the inner denticulated margin of P5 swimming legs and the shape of the mandibular cutting blade are morphological characters that have been described early in the literature as species-specific (Jaschnov 1955;

Beklemishev 1959; Frost 1974; Vyshkvartzeva 1976; Brodskii et al. 1983). However, due to the arduousness of their examination, they remain rarely used to identifyCalanusspecies.

Only 23 individuals out of the 71 examined exhibited the species-specific features typical for the species they belong to (verified by genetics), according to the literature (Supporting Information Table 3). For the other individuals, the morphological characteristics examined were different from that of the species according to the literature (Figs. 9, 10). Further- more, no geographic coherence was found in the deviations of the characteristics (Supporting Information Table 3). This resulted in an error rate of 30% and 31% in the identification decisions made by the experts in Calanusmorphology, after comparing their decision with results of genetic identification. Identification decisions of both experts matched only 36 times, and of these only 32 individuals (45% of the total) were confirmed to be correct by genetic identification.

More specifically, experts’ decision and genetics matched at Table 2. Pearson’s r calculation for testing the correlation

betweenCalanusbody size (prosome length) and latitude.

Species C. finmarchicus C. glacialis

Stage CIV CV CVI F CIV CV CVI F

n 21 161 92 201 269 151

Pearson’sr 0.8** 0.39** 0.19 0.74** 0.65** 0.84**

Significance levels (pvalue) are indicated by: *p<0.05 and **p<0.01.

(9)

51% for the individuals at stage CV, while experts’ decision and genetics only matched at 32% for the adult female individuals. It has to be kept in mind that the morphological features described in literature to discriminate betweenCala- nus species are typically described and can be applied directly for identification of adult females (or males) only, while we tested them on both adult females and CVs. They may not work for distinguishing copepodids at pre-adult CV stage, as some morphological structures are still not fully developed or expressed. However, the misidentification of 68% of adult females and disagreement between two experts is striking. In a few cases, the characteristics observed in

genetically identified species had appearance theoretically typical of the opposite species. Part of the problem may result from the fact that the characteristics are at the moment predominantly of a descriptive type and they have been portrayed based on “typical” individuals from only a few sites over the species distribution range.

To conclude, the morphological characters involving the 5^th pair of legs and the gnathobase were not consistent enough to be used for species identification. Therefore, we cannot recommend using these characteristics to reliably identify C. finmarchicus and C. glacialis without additional investigations.

Fig. 7.Red pigmentation onCalanus finmarchicus(C. fin) andCalanus glacialis(C. gla) antennules in different regions. Species-related redness from four regions where both species co-occur: the Nansen Basin, Svalbard, Saltenfjord/Skjerstadfjord, and Lurefjord; and from the White Sea where onlyC. glacialis occurs, and Raunefjord/Korsfjord where onlyC. finmarchicusoccurs. Blue color of the pie charts indicates proportion of individuals for which less than 10%

of the surface of their antennules was red; red color indicates proportion of individuals for which more than 10% of red pigmentation was noticed.

(10)

Discussion

Characters variability

The smaller size ofC. glacialisin the Norwegian fjord populations, compared to high Arctic populations, largely explains why the species wide boreal occurrence (Choquet et al. 2017) has not been detected before. For instance, occurrence of C. glacialis in the southern Lurefjord was not detected before molecular markers were applied (Bucklin et al. 2000). In the context of climate change and ocean warming, it is to be expected that more and moreC. glacialis individuals will be able to complete their life-cycle within a year, and then have a body size comparable to that ofC. finmarchicus. The decrease in body size with decreasing latitude is likely a direct effect of temperature (Atkinson and Sibly 1997), but variation in the duration of the productive season and predation pressure by visual predators (Brooks and Dod- son 1965) may also play an important role. Copepods are ectothermic, they primarily rely on external sources to regu- late their body heat. The temperature-size-rule refers to the widely observed phenomenon that ectotherms reared at lower temperatures usually grow more slowly, but become larger as adults compared to individuals reared at higher temperatures (Atkinson 1994; Atkinson and Sibly 1997). Cal- anoid copepods appear especially sensitive to temperature by having a fourfold greater reduction in adult body mass per degree Celsius compared to Cyclopoid copepods (Horne et al. 2016). Increasing latitude and mean temperature are strongly correlated (Sunday et al. 2011), and distinguishing separate effects may not be straight forward. However, oxygen demand and supply has been suggested as a driver of both processes (Horne et al. 2015), as the metabolic demand increases with increasing temperature, while the oxygen Fig. 8.Antennules redness frequency distribution perCalanusspecies.

This violin graph was realized under RStudio v.1.0.143 with the package ggplot2 (Wickham 2009). The graph shows the distribution of each species individuals on the following three ranks scale of redness: 15less than 10% of redness; 2510–50% redness; 35more than 50% redness.

Table 3. Antennules red pigmentation in copepodite stages CIV, CV, and adult females (CVI F) and males (CVI M) ofCala- nus finmarchicus(C. fin) andCalanus glacialis(C. gla) from different geographical locations.

Antennules redness

Species Stage <10% 10–50% 50–90% >90% Total a. Nansen Basin

C. fin CV 100% 0% 0% 0% 23

CVI F 88% 12% 0% 0% 24

Total 94% 6% 0% 0% 47

C. gla CVI F 0% 4% 85% 11% 47

b. Svalbard

C. fin CIV 100% 0% 0% 0% 1

CV 40% 20% 20% 20% 10

CVI F 59% 35% 6% 0% 49

CVI M 100% 0% 0% 0% 1

Total 58% 31% 8% 3% 61

C. gla CIV 0% 1% 9% 90% 98

CV 2% 4% 21% 73% 56

CVI F 8% 25% 33.5% 33.5% 12

Total 1% 4% 15% 80% 166

c. White Sea

C. gla CIV 93% 7% 0% 0% 100

CV 0% 57% 36% 7% 14

CVI F 0% 0% 100% 0% 1

Total 81% 13% 5% 1% 115

d. Salten/Skjerstadfj.

C. fin CIV 67% 0% 0% 33% 3

CV 100% 0% 0% 0% 33

CVI F 90% 10% 0% 0% 20

CVI M 100% 0% 0% 0% 4

Total 95% 3% 0% 2% 60

C. gla CIV 0% 0% 0% 1% 1

CV 20% 20% 38% 22% 40

CVI F 4% 16% 40% 40% 84

CVI M 100% 0% 0% 0% 5

Total 12% 16% 38% 34% 130

e. Lurefjord (in 3 categories)

C. fin CIV 0% 0% 100% 3

CV 95% 0 5% 22

Total 84% 0 16% 25

C. gla CIV 0 0 100% 1

CV 25% 0 75% 158

CVI F 100% 0 0 5

Total 27% 0 73% 164

f. Raune/Korsfj.

C. fin CIV 14% 8% 64% 14% 14

CV 63% 18% 18% 1% 74

Total 56% 16% 25% 3% 88

(11)

availability in the water decreases (Verberk et al. 2011). In addition, on-going climate change that is impacting the temperature ofCalanushabitat brings another unpredictable variable affecting body size ofCalanus species. Predation by visual predators, such as fish, may also induce a change in body-size composition in zooplankton communities. In the classical study by Brooks and Dodson (1965), the zooplankton community shifted from dominance of large- to dominance of small species in a freshwater lake after a fish- predator was introduced. According to optimal foraging the- ory, predators should target larger sized prey when handling time is a restriction. Both modeling studies and field investiga- tion confirm that lesser sandeel (Ammodytes marinus) in the North Sea actively target large copepods, such asC. finmarchicus, over smaller copepod taxa when these are available (van Deurs et al. 2014; van Deurs et al. 2015). On a longer time- scale, adaptive responses to predation pressure on the larger species may result in a dominance of species with shorter life- spans and smaller body-size (Stearns 1992; Berge et al. 2012).

However, to the best of our knowledge, there are no studies showing that predation may cause intraspecific changes in body size within populations ofCalanusspp.

It has been proposed that the pigment involved in redness of Calanoid copepods is astaxanthin, a form of keto- carotenoid (Mojib et al. 2014). This pigment has a role in the protection against UVR irradiance, and usually appears red in copepods. Copepods can adjust their level of astaxanthin pigment quickly, even within a season, depending on the prevailing threat, UVR, or predators (Hansson 2000).

Given such variability, it is thus not surprising that redness cannot be used reliably as a species diagnostic tool. Examina- tion of more samples for each developmental stage, from different depths, and seasonal observations may help to better understand the reasons for variability of red pigmentation in Calanusand its relation to environmental parameters.

Biological implications

Copepod species of the genus Calanusare the most studied among the zooplankton. They are often used as biological indicators of water masses and to follow the effects of

climate change on the marine ecosystems. However, in the majority of past studies, species identification has been based on morphometric and morphological characteristics. We found that none of the morphometric and morphological characteristics used in literature allow for unequivocal identification and separation of species. Therefore, it is likely that our knowledge of Calanus geographical distribution is plagued by species misidentification. Predictions on climate change effects and ecological models based on the present view of Calanus distribution and stocks dynamics are thus likely to be at least partially erroneous, especially in the areas of sympatry. Furthermore, as Calanus species distributions are expected to change and overlap even more in response to global warming (Slagstad et al. 2011), the systematic use of molecular identification is required to document these changes. Considering the importance of Calanusrange shifts Table 4. Redness of C. finmarchicus and C. glacialis female

genital somite. Any red pigmentation observed on one or both spermathecae (5 genital field) was reported as “Red”. No redness at all was reported as “Pale.”

Species C. finmarchicus C. glacialis Genital somite Red Pale Total Red Pale Total

Svalbard 0 100% 49 73% 27% 11

White Sea 0 0 0 100% 0 1

Salten/Skjerstadfj. 0 100% 20 90% 10% 82

Lurefjord 0 0 0 80% 20% 5

Total 0 100% 69 88% 12% 99

Table 5. Redness of antennules vs. genital somite for C. finmarchicus andC. glacialisfemales in distinct locations. Numbers of individuals with different combinations of pigmentation (antennules vs. genital somite) are reported. For the antennules, every individual with>10% redness was considered “Red”. For the genital somite, every individual with any distinguishable red pigmentation was considered “Red.”

Redness Svalbard

Antennules Genital somite C. finmarchicus C. glacialis

Pale Pale 19 1

Red Red 0 8

Pale Red 0 0

Red Pale 20 2

Redness White Sea

Pale Pale 0 0

Red Red 0 1

Pale Red 0 0

Red Pale 0 0

Redness Salten/Skjerstadfj.

Pale Pale 18 0

Red Red 0 71

Pale Red 0 3

Red Pale 2 8

Redness Lurefjord

Pale Pale 0 1

Red Red 0 0

Pale Red 0 4

Red Pale 0 0

(12)

for our understanding of climate change impact on pelagic ecosystems (Beaugrand and Kirby 2010; Søreide et al. 2010), it is critical to tease apart the respective effects of

morphological misidentification from on-going range shifts and this will require a thorough reassessment of historical distribution using molecular tools.

Fig. 9.Morphology of the fifth thoracic pair of legs and the gnathobase of an adult femaleC. finmarchicusexhibiting traits theoretically assigned to C. glacialis. The specimen from Saltenfjord (ID: SALT27) exhibits concave denticulated lamellae with a well-expressed curvature, and a wide basis of the second ventral tooth, typical ofC. glacialisaccording to the literature.

Fig. 10.Morphology of the fifth thoracic pair of legs and the gnathobase of an adult femaleC. glacialisexhibiting traits theoretically assigned toC.

finmarchicus. The specimen from Saltenfjord (ID: SALT14) exhibits straight shaped denticulated lamellae, and a small second ventral tooth on the cutting edge of the coxa, typical ofC. finmarchicusaccording to the literature.

(13)

Misidentification ofCalanusspecies is also detrimental for our understanding of marine ecosystems. For example, a large part of the distribution range of C. glacialis has only been recently identified along the Norwegian coast (Choquet et al. 2017) questioning our understanding of fjords dynamics. In other words, life cycles, phenology and the exact role of each species within fjord ecosystems, potential for adapt- ability/resilience to climate variability, as well as response to environmental variations and population dynamics are not fully understood.

Comments and recommendations

None of the morphological characters described in the literature and re-assessed in the present study can reliably identify C. finmarchicus andC. glacialis with 100% confidence. There are some global trends that can bring information about the species composition though, but certainly not equally every- where. Prosome length may be useful to approximate the species composition in the Nansen Basin and in Svalbard waters, and likely in the Arctic Ocean. However, it is critical to keep in mind the underestimation ofC. glacialis. In fjords along the Norwegian coast, prosome length is clearly not usable, as the size range of both species overlaps completely. Regarding the redness of antennules/genital somite ofCalanus, it seems to be a useful indicator of species in the Arctic and in relatively closed fjords (with a sill—e.g., Saltenfjord, Skjerstadfjord, Lur- efjord), but not in open fjords (without sill—e.g., Raunefjord, Korsfjord). Again, by using this character, it is critical to keep in mind the variable error rates associated (Fig. 7). However, our data suggest that individuals with both red antennules and red genital somite can be identified asC. glacialis(but the opposite is not true, leading to an underestimation ofC. glacialis). We recommend not using the curvature of the inner denticulated margin of the P5 swimming legs and the shape of the mandibular cutting blade to discriminate between species, until the variability of these characters in all parts of the species distribution range is thoroughly investigated simultaneously with molecular identification.

The use of molecular tools is thus the only reliable method for discriminating between the two species. It is likely that the problems of identification encountered withCalanusalso exist in other taxa in pelagic zooplankton (e.g., Aarbakke et al. 2011).

Therefore, it is critical to start using molecular tools routinely for reliable species identification, especially for ecologically important organisms such asCalanus. Equipment, time, compe- tences needed, and cost related to molecular identification of Calanusare today a much lesser issue than it used to be. Indeed, as described in Smolina et al. (2014), the set of InDels markers that we used in the present study can be run on agarose gels and therefore used in a low-cost setting on board a research ves- sel. We also simplified the method of DNA extraction, which now consists of only removing the antennules of each individuals and incubating them 30 min in a buffer at no cost. With

these simplifications, genotyping 96 individuals ofCalanuscan be done in 5 h for less than 2 USD per individual.

References

Aarbakke, O. N. S., A. Bucklin, C. Halsband, and F. Norrbin.

2011. Discovery ofPseudocalanus moultoni(Frost, 1989) in Northeast Atlantic waters based on mitochondrial COI sequence variation. J. Plankton Res. 33: 1487–1495. doi:

10.1093/plankt/fbr057

Aßmus, J., W. Melle, D. Tjøstheim, and M. Edwards. 2009.

Seasonal cycles and long-term trends of plankton in shelf and oceanic habitats of the Norwegian Sea in relation to environmental variables. Deep-Sea Res. Part II Top. Stud.

Oceanogr.56: 1895–1909. doi:10.1016/j.dsr2.2008.11.004 Arnkværn, G., M. Daase, and K. Eiane. 2005. Dynamics of

coexistingCalanus finmarchicus,Calanus glacialisandCala- nus hyperboreus populations in a high-Arctic fjord. Polar Biol.28: 528–538. doi:10.1007/s00300-005-0715-8

Atkinson, D. 1994. Temperature and organism size - a biological law for ectotherms, p. 1–58. In M. Begon and A.

H. Fitter [eds.], Advances in ecological research, v. 25.

Academic Press.

Atkinson, D., and R. M. Sibly. 1997. Why are organisms usually bigger in colder environments? Making sense of a life history puzzle. Trends Ecol. Evol. 12: 235–239. doi:

10.1016/S0169-5347(97)01058-6

Beaugrand, G., K. M. Brander, J. A. Lindley, S. Souissi, and P.

C. Reid. 2003. Plankton effect on cod recruitment in the North Sea. Nature 426: 661–664. doi:10.1038/

nature02164

Beaugrand, G., and R. R. Kirby. 2010. Climate, plankton and cod. Glob. Chang. Biol. 16: 1268–1280. doi:10.1111/

j.1365-2486.2009.02063.x

Beklemishev, K. V. 1959. On the anatomy of masticatory organs of Copepoda. Report 2: The masticatory edge in mandibles of certain species of Calanidae and Eucalani- dae. Trudy Inst. Okeanol.30: 148–155.

Berge, J., T. M. Gabrielsen, M. Moline, and P. E. Renaud.

2012. Evolution of the Arctic Calanuscomplex: An Arctic marine avocado? J. Plankton Res. 34: 191–195. doi:

10.1093/plankt/fbr103

Blachowiak-Samolyk, K. 2008. Contrasting zooplankton communities (Arctic vs. Atlantic) in the European Arctic marginal ice zone. Oceanologia50: 363–389.

Bonnet, D., and C. Frid. 2004. Seven copepod species considered as indicators of water-mass influence and changes:

Results from a Northumberland coastal station. ICES J.

Mar. Sci.61: 485–491. doi:10.1016/j.icesjms.2004.03.005 Brodskii, K., N. Vyshkvartseva, M. Kos, and E. Markhatseva.

1983. Copepods (Copepoda: Calanoida) of the seas of the USSR and adjacent waters, p. 358 [in Russian]. Keys to the fauna of the USSR, v. 1.

(14)

Broms, C., W. Melle, and S. Kaartvedt. 2009. Oceanic distribution and life cycle of Calanusspecies in the Norwegian Sea and adjacent waters. Deep-Sea Res. Part II Top. Stud.

Oceanogr.56: 1910–1921. doi:10.1016/j.dsr2.2008.11.005 Brooks, J. L., and S. I. Dodson. 1965. Predation, body size,

and composition of plankton. Science 150: 28–35. doi:

10.1126/science.150.3692.28

Brun, P., M. R. Payne, and T. Kiørboe. 2016. Trait biogeogra- phy of marine copepods – an analysis across scales. Ecol.

Lett.19: 1403–1413. doi:10.1111/ele.12688

Bucklin, A., S. Kaartvedt, M. Guarnieri, and U. Goswami.

2000. Population genetics of drifting (Calanus spp.) and resident (Acartia clausi) plankton in Norwegian fjords. J.

Plankton Res. 22: 1237–1251. doi:10.1093/plankt/

22.7.1237

Choquet, M., and others. 2017. Genetics redraws pelagic bio- geography of Calanus. Biol. Lett. 13: 588. doi:10.1098/

rsbl.2017.0588

Conover, R. 1988. Comparative life histories in the genera Calanusand Neocalanusin high latitudes of the northern hemisphere, p. 127–142. In G. A. Boxshall and H. K.

Schminke [eds.], Biology of copepods. Springer.

Daase, M., and K. Eiane. 2007. Mesozooplankton distribution in northern Svalbard waters in relation to hydrography.

Polar Biol.30: 969–981. doi:10.1007/s00300-007-0255-5 Daase, M., and others. 2013. Timing of reproductive events

in the marine copepodCalanus glacialis: A pan-Arctic per- spective. Can. J. Fish. Aquat. Sci. 70: 871–884. doi:

10.1139/cjfas-2012-0401

Falk-Petersen, S., V. Pavlov, S. Timofeev, and J. R. Sargent.

2007. Climate variability and possible effects on arctic food chains: The role of Calanus, p. 147–166. In J. B.

Ørbaek, R. Kallenborn, E. N. Hegseth, S. Falk-Petersen, and A. H. Hoel [eds.], Arctic alpine ecosystems and people in a changing environment. Springer.

Falk-Petersen, S., P. Mayzaud, G. Kattner, and J. Sargent.

2009. Lipids and life strategy of ArcticCalanus. Mar. Biol.

Res.5: 18–39. doi:10.1080/17451000802512267

Fleminger, A., and K. Hulsemann. 1977. Geographical range and taxonomic divergence in North Atlantic Calanus (C.

helgolandicus, C. finmarchicus and C. glacialis). Mar. Biol.

40: 233–248. doi:10.1007/BF00390879

Forest, A., and others. 2011. Biogenic carbon flows through the planktonic food web of the Amundsen Gulf (Arctic Ocean): A synthesis of field measurements and inverse modeling analyses. Prog. Oceanogr. 91: 410–436. doi:

10.1016/j.pocean.2011.05.002

Frost, B. 1974. Calanus marshallae, a new species of calanoid copepod closely allied to the sibling speciesC. finmarchicus and C. glacialis. Mar. Biol. 26: 77–99. doi:10.1007/

BF00389089

Gabrielsen, T. M., and others. 2012. Potential misidentifica- tions of two climate indicator species of the marine arctic

ecosystem: Calanus glacialis and C. finmarchicus. Polar Biol.35: 1621–1628. doi:10.1007/s00300-012-1202-7 Gislason, A., and O. S. Astthorsson. 2002. The food of Nor-

wegian spring-spawning herring in the western Norwe- gian Sea in relation to the annual cycle of zooplankton.

Sarsia87: 236–247. doi:10.1080/00364820260294860 Hansson, L. A. 2000. Induced pigmentation in zooplankton:

A trade–off between threats from predation and ultravio- let radiation. Proc. R. Soc. Lond. B Biol. Sci. 267: 2327–

2331. doi:10.1098/rspb.2000.1287

Hassel, A. 1986. Seasonal changes in zooplankton composition in the Barents Sea, with special attention toCalanus spp.(Copepoda). J. Plankton Res.8: 329–339. doi:10.1093/

plankt/8.2.329

Hays, G. C., A. J. Richardson, and C. Robinson. 2005. Cli- mate change and marine plankton. Trends Ecol. Evol.20:

337–344. doi:10.1016/j.tree.2005.03.004

Helaou€et, P., and G. Beaugrand. 2007. Macroecology ofCala- nus finmarchicusandC. helgolandicusin the North Atlantic Ocean and adjacent seas. Mar. Ecol. Prog. Ser. 345: 147–

165. doi:10.3354/meps06775

Hirche, H.-J., W. Hagen, N. Mumm, and C. Richter. 1994.

The northeast water polynya, Greenland Sea. Polar Biol.

14: 491–503. doi:10.1007/BF00239054

Hirche, H.-J., and K. Kosobokova. 2011. Winter studies on zooplankton in Arctic seas: The Storfjord (Svalbard) and adjacent ice-covered Barents Sea. Mar. Biol. 158: 2359.

doi:10.1007/s00227-011-1740-5

Horne, C. R., A. G. Hirst, and D. Atkinson. 2015. Tempera- ture-size responses match latitudinal-size clines in arthro- pods, revealing critical differences between aquatic and terrestrial species. Ecol. Lett. 18: 327–335. doi:10.1111/

ele.12413

Horne, C. R., A. G. Hirst, D. Atkinson, A. Neves, and T.

Kiørboe. 2016. A global synthesis of seasonal temperature- size responses in copepods. Glob. Ecol. Biogeogr.25: 988–

999. doi:10.1111/geb.12460

Huys, R., and G. A. Boxshall. 1991. Copepod evolution. Ray Society.

Jaschnov, W. A. 1955. Morphology, distribution and system- atics of Calanus finmarchicussl [in Russian]. Zool. Zh.34:

1210–1223.

Jaschnov, W. A. 1966. Water mass and plankton: 4.Calanus finmarchicus andDimophyses arcticaas indicators of Atlan- tic waters in the Polar Basin. Oceanol. Acad. Sci. USSR 6:

404–412.

Jaschnov, W. A. 1970. Distribution ofCalanus species in the seas of the northern hemisphere. Int. Revue ges. Hydro- biol. Hydrogr.55: 197–212. doi:10.1002/iroh.19700550203 Jaschnov, W. A. 1972. On the systematic status of Calanus

glacialis, Calanus finmarchicus and Calanus helgolandicus.

Crustaceana22: 279–284. doi:10.1163/156854072X00561

(15)

Kosobokova, K. 2012. Zooplankton of the Arctic Ocean:

Community structure, ecology, spatial distribution, p. 272 [in Russian]. GEOS.

Kosobokova, K., and H.-J. Hirche. 2009. Biomass of zooplankton in the eastern Arctic Ocean–a base line study. Prog. Oce- anogr.82: 265–280. doi:10.1016/j.pocean.2009.07.006 Kosobokova, K. N., R. R. Hopcroft, and H.-J. Hirche. 2011.

Patterns of zooplankton diversity through the depths of the Arctic’s central basins. Mar. Biodivers.41: 29–50. doi:

10.1007/s12526-010-0057-9

Kwasniewski, S., H. Hop, S. Falk-Petersen, and G. Pedersen.

2003. Distribution of Calanus species in Kongsfjorden, a glacial fjord in Svalbard. J. Plankton Res. 25: 1–20. doi:

10.1093/plankt/25.1.1

Lindeque, P., R. Harris, M. Jones, and G. Smerdon. 1999.

Simple molecular method to distinguish the identity of Calanus species (Copepoda: Calanoida) at any developmental stage. Mar. Biol. 133: 91–96. doi:10.1007/

s002270050446

Lindeque, P. K., S. J. Hay, M. R. Heath, A. Ingvarsdottir, J.

Rasmussen, G. R. Smerdon, and J. J. Waniek. 2006. Inte- grating conventional microscopy and molecular analysis to analyse the abundance and distribution of fourCalanus congeners in the North Atlantic. J. Plankton Res.28: 221–

238. doi:10.1093/plankt/fbi115

Michaud, J., and C. T. Taggart. 2007. Lipid and gross energy content of North Atlantic right whale food, Calanus finmarchicus, in the Bay of Fundy. Endanger. Species Res. 3:

77–94. doi:10.3354/esr003077

Mojib, N., M. Amad, M. Thimma, N. Aldanondo, M.

Kumaran, and X. Irigoien. 2014. Carotenoid metabolic profiling and transcriptome-genome mining reveal functional equivalence among blue-pigmented copepods and appendicularia. Mol. Ecol. 23: 2740–2756. doi:10.1111/

mec.12781

Montero-Pau, J., A. Gomez, and J. Mu~noz. 2008. Application of an inexpensive and high-throughput genomic DNA extraction method for the molecular ecology of zooplank- tonic diapausing eggs. Limnol. Oceanogr.: Methods 6:

218–222. doi:10.4319/lom.2008.6.218

Nielsen, T. G., S. Kjellerup, I. Smolina, G. Hoarau, and P.

Lindeque. 2014. Live discrimination of Calanus glacialis and C. finmarchicus females: Can we trust phenological differences? Mar. Biol. 161: 1299–1306. doi:10.1007/

s00227-014-2419-5

Parent, G. J., S. Plourde, and J. Turgeon. 2011. Overlapping size ranges of Calanus spp. off the Canadian Arctic and Atlantic Coasts: Impact on species’ abundances. J. Plank- ton Res.33: 1654–1665. doi:10.1093/plankt/fbr072 Skjoldal, H. R. 2004. An introduction to the Norwegian Sea

ecosystem, p. 15–32.InH. R. Skjoldal, R. Sætre, A. Fernø, O. A. Misund, and I. Røttingen [eds.], The Norwegian Sea ecosystem. Tapir Academic Press.

Slagstad, D., I. H. Ellingsen, and P. Wassman. 2011. Evaluat- ing primary and secondary production in an Arctic Ocean void of summer sea ice: An experimental simulation approach. Prog. Oceanogr. 90: 117–131. doi:10.1016/

j.pocean.2011.02.009

Smolina, I., and others. 2014. Genome- and transcriptome- assisted development of nuclear insertion/deletion markers forCalanusspecies (Copepoda: Calanoida) identification. Mol. Ecol. Resour. 14: 1072–1079. doi:10.1111/

1755-0998.12241

Søreide, J. E., S. Falk-Petersen, E. N. Hegseth, H. Hop, M. L.

Carroll, K. A. Hobson, and K. Blachowiak-Samolyk. 2008.

Seasonal feeding strategies of Calanus in the high-Arctic Svalbard region. Deep-Sea Res. Part II Top. Stud. Ocean- ogr.55: 2225–2244. doi:10.1016/j.dsr2.2008.05.024 Søreide, J. E., E. Leu, J. Berge, M. Graeve, and S. Falk-

Petersen. 2010. Timing of blooms, algal food quality and Calanus glacialis reproduction and growth in a changing Arctic. Glob. Chang. Biol. 16: 3154–3163. doi:10.1111/

j.1365-2486.2010.02175.x

Stearns, S. C. 1992. The evolution of life histories Oxford Univ. Press.

Steen, H., D. Vogedes, F. Broms, S. Falk-Petersen, and J.

Berge. 2007. Little auks (Alle alle) breeding in a High Arc- tic fjord system: Bimodal foraging strategies as a response to poor food quality? Polar Res. 26: 118–125. doi:

10.1111/j.1751-8369.2007.00022.x

Sunday, J. M., A. E. Bates, and N. K. Dulvy. 2011. Global analysis of thermal tolerance and latitude in ectotherms.

Proc. R. Soc. B Biol. Sci. 278: 1823–1830. doi:10.1098/

rspb.2010.1295

Unstad, K. H., and K. S. Tande. 1991. Depth distribution of Calanus finmarchicus and C. glacialis in relation to environmental conditions in the Barents Sea. Polar Res. 10:

409–420. doi:10.3402/polar.v10i2.6755

van Deurs, M., M. Koski, and A. Rindorf. 2014. Does copepod size determine food consumption of particulate feeding fish? ICES J. Mar. Sci.71: 35–43. doi:10.1093/icesjms/fst090 van Deurs, M., C. Jorgensen, and O. Fiksen. 2015. Effects of copepod size on fish growth: A model based on data for North Sea sandeel. Mar. Ecol. Prog. Ser. 520: 235–243.

doi:10.3354/meps11092

Varpe, Ø., Ø. Fiksen, and A. Slotte. 2005. Meta-ecosystems and biological energy transport from ocean to coast: The ecological importance of herring migration. Oecologia 146: 443. doi:10.1007/s00442-005-0219-9

Verberk, W., D. T. Bilton, P. Calosi, and J. I. Spicer. 2011.

Oxygen supply in aquatic ectotherms: Partial pressure and solubility together explain biodiversity and size patterns.

Ecology92: 1565–1572. doi:10.1890/10-2369.1

Vyshkvartzeva, N. V. 1972. The structure of mandibles in Copepoda (the genusCalanus) in relation to the latitudinal zonality. Issled. Fauny Morei12: 161–171.

(16)

Vyshkvartzeva, N. V. 1976. The functional morphology of mouth parts of the species Calanus sl (Copepoda, Cala- noida). Issled. Fauny Morei18: 11–69.

Wassmann, P., and others. 2006. Food webs and carbon flux in the Barents Sea. Prog. Oceanogr. 71: 232–287. doi:

10.1016/j.pocean.2006.10.003

Wassmann, P., and others. 2015. The contiguous domains of Arctic Ocean advection: Trails of life and death. Prog.

Oceanogr.139: 42–65. doi:10.1016/j.pocean.2015.06.011 Wickham, H. 2009. ggplot2: Elegant graphics for data analy-

sis. Springer-Verlag.

Wilson, S. E., D. K. Steinberg, and K. O. Buesseler. 2008.

Changes in fecal pellet characteristics with depth as indicators of zooplankton repackaging of particles in the mesopelagic zone of the subtropical and subarctic North Pacific Ocean. Deep-Sea Res. Part II Top. Stud. Oceanogr.

55: 1636–1647. doi:10.1016/j.dsr2.2008.04.019

Acknowledgments

We are very grateful to Morten Krogstad and to the crew and captain of the R/VG.O. Sarsfor their help with the sampling. We thank Mikko

Vihtakari for his suggestions and practical help on treatment of pigmentation data and statistics. We also thank the research network ARCTOS for useful collaborations. The project was funded by the European Com- mission FP7 EURO-BASIN (Grant agreement: 264 933), the Norwegian Research Council (project HAVKYST 216578 and PolarProg 227139 and 246747) and Nord University. M. C. was supported by a Ph.D. student fellowship from Nord University, and M. H. by a Ph.D. student fellowship from The University Centre in Svalbard. K. K. was supported by Rus- sian Foundation for Basic Research (15-29-02447; 16-04-00375) and Russian Scientific Foundation (14-50-00095). M. D. was supported by NRC Grant 226417 Marine Night. S. K. was supported by the Polish- Norwegian Research Program Pol-Nor/201992/93/2014 (Project DWARF).

Conflict of Interest None declared.

Submitted 06 October 2017 Revised 18 December 2017 Accepted 08 January 2018

Associate editor: Malinda Sutor